The experiments described in this proposal investigate the process by which young children learn new words. Children as young as two years old can rapidly assign a meaning to a new word that they hear. There is a strong intuitive assumption that to establish this meaning, learners must hear the word in the presence of its referent, for example, hearing [kaet] in the presence of a cat, or [kIk] in the presence of a kickig action. Yet, in the natural course of events, words, especially verbs, are often introduced in the absence of their referents (e.g., Let's pack our bag). Strikingly, analyses reveal that when conversing with their young children, most of the verbs produced by mothers refer to absent events. How, then, do they establish meaning for new verbs? We aim to develop a paradigm and offer an empirical foundation to address this issue. We do so by presenting 2-year-olds with novel verbs in linguistic contexts only, without a relevant visual scene. After a delay, we offer them candidate visual events to determine if they have (a) assigned meaning to the novel verbs, and (b) can recall these meanings even after a delay. We also assess whether a period of sleep during the delay helps toddlers integrate the new verb into their existing lexical knowledge. The proposed work aims to characterize the process of vocabulary acquisition. Understanding the mechanisms of language acquisition is crucial not only to a theoretical understanding of language acquisition and cognitive development, but also has implications for helping children with developmental delays. The research described here contributes to a growing base of knowledge about learning in typically-developing children, which is key in understanding how learning may diverge in children with language delays and disabilities.